The goal of our research is to define the initiation mechanism of gene amplification in cancer cells. Defining the initiation mechanism should provide important knowledge, as amplification very often drives tumor progression and therapy resistance. An important chromosomal structure at a very early step of gene amplification is an Inverted duplication of large chromosomal regions. This indicates a following process (breakage-fusion-bridge cycle, BFB cycle) as a potential initiation mechanism; (1) a chromosome break leads to the fusion of the broken ends after replication (sister chromatid fusion), resulting in a de novo palindromic (inverted duplication) chromosome with two centromeres; (2) subsequent mitosis causes a tension between the two centromeres (bridge), resulting in a partially duplicated chromosome with a broken end (break); and (3) the broken end enters into another cycle. Therefore, the initial step is a potential rate-limiting step, because, once a palindromic dicentric chromosome is generated, it would inevitably enters into a bad spiral (BFB cycle) and leads to the accumulation of specific genomic regions (palindromic gene amplification). In this proposal, we will define cis- (genomic) and trans- (genetic) acting factors that are important in the initiation of palindromic gene amplification in human tumors. We have previously shown in mammalian cell models that a DNA inverted repeat (DNA-IR) pre-existing in the genome, with an adjacent DSB, greatly promotes palindromic gene amplification. Thus, an initial step of gene amplification is DSB-initiated, illegitimate recombination between the repeats of a DNA-IR. This leads to our DNA-level model; fold-back (Intra-strand) annealing within a DNA-IR would generate a chromosome with a hairpin-capped end, and subsequent DNA replication would complete palindromic duplication. Based on our model, we will test potential determinants for palindromic gene amplification: pre-existing DNA-IRs in the human genome are cis- acting (genomic) determinant (Aim 1), and genes that process a hairpin-capped end and physiological conditions that induce DSBs are important trans-acting determinants (Aim 2). To accomplish our aims, we have established unique experimental systems that can overcome the difficulties in studying palindromic DNA. The genomic approach employs a novel technique for the enrichment of palindromic DNA and identifies important DNA-IRs for palindromic gene amplification in primary human tumors. Our cell culture system is designed to measure the occurrence of DSB-initiated illegitimate recombination at a DNA-IR in cells with a variety of genetic backgrounds. These experiments should collectively identify specific DNA-IRs as an At-Risk Motif for developing gene amplification. Given the impact of structural variations in the normal human genome, polymorphic DNA-IRs could be an important predictor of an individual's susceptibility to gene amplification.